Ordering Information:
ORDER NUMBER:  91212
DATE AVAILABLE: Spring 2008

Printed Report		PDF
Subscribers	Order Report
Non-Subscriber - Coming August / Nov. 08	AWWA Bookstore	N/A
	IWA Publishing	N/A

PRINCIPAL INVESTIGATORS:
Jörg E. Drewes, Christopher Bellona, Pei Xu, Gary L. Amy, Gerry Filteau, and Gregg Oelker

OBJECTIVES:
This purpose of this study was to explore whether nanofiltration (NF) and ultra-low pressure reverse osmosis (ULPRO) membranes can consistently meet potable water quality requirements with respect to total organic carbon (TOC), total nitrogen, and regulated and unregulated trace organic compounds. The goals were also to determine whether or not operating characteristics of NF and ULPRO membranes (such as flux, fouling/scaling, and cleaning frequencies) are comparable to conventional thin-film composite RO membranes and operating feed pressure requirements are significantly lower than conventional RO.

BACKGROUND:
Membrane treatment of source water of impaired quality by an integrated membrane system (IMS), such as microfiltration (MF) pretreatment followed by RO, represents the industry standard for drinking water augmentation projects. An alternative IMS involving NF membranes and ULPRO membranes in place of conventional RO membranes provides an opportunity for lower pressure/higher flux operating conditions and higher selectivity (e.g., targeting trace organics over monovalent salts).

HIGHLIGHTS:
The selection of NF and ULPRO membranes for pilot-scale testing became an exercise in balancing the benefits of lower pressure and energy requirements with constituent rejection and permeate quality. Rejection of trace organics, however, improved with membrane fouling, which is especially important for NF membranes, because initial (organic) fouling will always occur where membranes are employed to treat surface water or secondary and tertiary treated effluents. The transport of hydrophilic ionic organic contaminants was especially hindered in NF membranes as a result of improved electrostatic exclusion, pore clogging, and an increased adsorption capacity of fouled polyamide membranes.

APPROACH:
This work involved the development and validation of a laboratory-scale membrane testing protocol to select viable membranes for pilot- and full-scale operation. This selection protocol balanced operational characteristics with product water quality and allowed for a pre-selection of potentially viable candidate membranes. Membranes considered for selection were characterized as thin-film composite polyamide membranes and included commercially available ULPRO and NF membrane products. Three candidate membranes were selected and each tested using a 70 L/min (18 gpm) membrane pilot skid for at least 1,300 hours on microfiltered feed water at two full-scale facilities. Findings of this study were compiled into a model framework to describe and predict the rejection of organic micropollutants during NF or RO treatment.

RESULTS/FINDINGS:
Membranes tested during this study using the membrane testing protocol displayed a wide range of operating characteristics including fouling tendency, permeate productivity, and feed pressure while displaying a wide range of rejection behavior in terms of TOC, ammonium, nitrate, and select trace organic contaminant removal. The TMG10 (Toray America) and NF-90 (Dow/Filmtec) membranes were selected as candidate membranes for pilot-scale testing based upon high specific fluxes after a period of fouling as compared to the benchmark RO membrane (TFC-HR, Koch Membrane Systems) and efficient rejection of TOC (<0.5 mg/L), nitrogen species (<10 mg-N/L), and organic trace contaminants. Although the membrane NF-4040 (Dow/Filmtec) did not meet the nitrogen and TOC water quality criteria, this membrane was also selected for pilot-scale testing due to moderate flux decline, efficient trace organic rejection (usually greater than 90 percent), and a significantly higher specific flux than the benchmark TFC-HR membrane. Rejection of regulated organics was found to be similar to the benchmark TFC-HR membrane with a poor to moderate rejection of chloroform and N-nitrosodimethylamine (NDMA). All three candidate membranes were highly efficient in rejecting unregulated organics with non-detectable concentrations or only traces of the target compounds in the permeates.

IMPACT:
ULPRO membranes can consistently meet potable water quality requirements for treating impaired source waters with respect to TOC, total nitrogen, and both regulated and unregulated trace organic compounds. ULPRO membranes exhibit specific flux values that were slightly higher as compared to conventional RO membranes. Because purchase cost and expected life of these membranes are similar, ULPRO membranes potentially offer lower operating expenses and therefore should be considered in membrane procurement or replacement programs of an IMS. Achieving significant cost savings associated with lower feed pressure, however, is limited to certain low fouling loose NF membranes. While these membranes demonstrated efficient rejection of TOC and a high selectivity for a wide range of trace organics, they are limited in rejecting ammonia and nitrate.

RESEARCH PARTNERS:

• WateReuse Foundation
• West Basin Municipal Water District, Carson, Calif.

PARTICIPANTS:
City of Scottsdale, Ariz.